Considerations of prescribers and pharmacists for the use of non‐selective β‐blockers in asthma and COPD patients: An explorative study

Abstract Rationale, aims, and objectives Despite recommendations in prevailing guidelines to avoid the use of non‐selective (NS) β‐blockers in patients with asthma or COPD, on average, 10 patients per community pharmacy receive NS β‐blockers monthly. The aim of our study was to identify the reasons of prescribers and pharmacists to treat asthma and COPD patients with NS β‐blockers. Methods Fifty‐three community pharmacists in the Netherlands selected patients with actual concurrent use of inhalation medication and NS β‐blockers. For at least 5 patients, each pharmacist screened all medication surveillance signals and actions taken at first dispensing. Each pharmacist selected 3 different initial prescribers for a short interview to explore their awareness of the co‐morbidity and reasons to apply NS β‐blockers. Results Pharmacists identified 827 asthma/COPD patients with actual use of NS β‐blockers. From these, 153 NS β‐blocker prescribers were selected and interviewed (64 general practitioners, 45 ophthalmologists, 24 cardiologists, and 20 other prescribers). One hundred seven prescribers were aware of the drug‐disease interaction of the asthma or COPD co‐morbidity when initiating the NS β‐blocker, and 46 were not. From these, 40 prescribers did not consider the contraindication to be relevant. For 299 patients, medication surveillance signals and actions at first dispensing were retrieved. Patients used predominantly ocular timolol (39.8%), and the oral preparations propranolol (30.8%) and carvedilol (15.1%). In 154 cases, the pharmacy system generated a warning alert. Conclusions A substantial number of prescribers was unaware of the co‐morbidity or did not regard NS β‐blockers contraindicated, despite prevailing clinical guidelines. Improvement programs should target prescribers' awareness and knowledge of NS β‐blockers in patients with asthma or COPD.


| Study design
This was an explorative observational study in 53 community pharmacies in the Netherlands. The study protocol was approved by the Ethical Committee of the Radboud UMC Nijmegen (approval number: 2015-2185).

| Setting
Fifty-three community pharmacists, located in different areas in the Netherlands, participated in this study between February and July 2016. These pharmacists belonged to 3 different educational groups within the national postgraduate specialization programme to become community pharmacists. Participating in research practice is part of the last year of this educational programme. Thus, the participating pharmacists were a convenience sample of all pharmacists in the Netherlands. Dutch pharmacists have a professional and legal responsibility for the drug treatment of their patients. 28 All pharmacists in the Netherlands use a computer system, designed to signalize drug-drug interactions and, if applicable, intolerances and drug-disease interactions. Pharmacists use these signals to identify drug therapy-related problems and consult prescribers.
Handling of all monitoring signals is registered in the local pharmacy system.

| Patient selection-identification
Patients were selected from pharmacy dispensing data. In the Netherlands, all medication dispensed on prescription is registered in the local pharmacy system. Dispensing data from more than 90% of the 1981 community pharmacies in the Netherlands are delivered routinely to the Foundation of Pharmaceutical Statistics (SFK). These data provide detailed information on the drugs dispensed, including the codes from the Anatomic Therapeutic Chemical (ATC) system of the World Health Organization. 29 The computerized pharmacy system can only calculate correct periods of drug use, if the total number of dispensed drug doses and the prescribed daily dose is entered. 30 From these data, SFK generates periodically online reports for participating pharmacies, to detect possible medication problems and to improve pharmaceutical care.
Participating pharmacists were provided with an automated web report that identified all users of inhalation medication for their pharmacy with actual use of NS (oral or ocular) β-blockers. The SFK web report presented all current users of inhalation medication for asthma or COPD (SABA, LABA, ICS or combinations of ICS/LABA, and short-and long-acting muscarinic antagonists; ATC code R03) that were also current users of an oral NS β-blocker (ATC codes C07AA, C07AG, C07BA, C07CA, C07DA, C07EA, and C07FA) or an ocular NS β-blocker (ATC codes: S01ED, except S01ED02 and S01ED52), independent of which medication had been started first.

| Selection of NS β-blocker initiators for prescriber interview
From the SFK web report, each pharmacist selected 3 prescribers from different disciplines for an interview about their choice to initiate the NS β-blocker for this population, based on a semistructured format (Appendix A). Pharmacists were asked to look for the first dispensing of an NS β-blocker in the patients dispensing history to identify the initial prescriber (initiator). A first dispensing was defined as the dispensing of an NS β-blocker without any dispensing of an NS β-blocker within the preceding 12 months. The first question in these interviews was whether the initiator was aware of the airway disease when starting the NS β-blocker. When this was the case, the pharmacist asked for the reasons to prescribe an NS β-blocker despite the drugdisease interaction and whether the choice would have been reconsidered if the patient would suffer from exacerbations after the start of the NS β-blocker. If the initiator was not aware, possible reasons for this were explored. Several possible reasons (eg, lack of a complete patient file and not interested in co-morbidities) were prepared in a digital form, with the possibility to add additional answers. and reasons for drug prescribing are often not communicated to the community pharmacist. As an approach, the pharmacy information system can generate "deduced contraindication" signals from the medication dispensed. These can be judged by the pharmacist, verified by the prescriber, and stored in the system. For example, when a patient uses antidiabetic medication, the pharmacist can enter the contraindication "diabetes mellitus" into the pharmacy information system.

| Analysis
With descriptive analysis, answers were stratified for medical specialism (GP, cardiologist or ophthalmologist, pulmonologist, neurologists, psychiatrist, and other specialists) and awareness of the contraindication, and categorized for reasons that the NS β-blocker was prescribed in spite of the airway disease. Differences between prescribers were examined with the chi-square test, using IBM Corp SPSS statistics, Chicago IL, USA, version 22.

| RESULTS
In 53 pharmacies, 827 patients were identified by the web report, dispensed NS β-blocker co-medication concomitantly with inhalation therapy for overlapping time periods. From this selection, medication surveillance signals were checked for 299 patients (Table 1) and 153 prescribers were interviewed.   In 122 cases (40.8%), the pharmacy information system did not generate any medication surveillance signal. In 154 cases, the system generated a contraindication signal (n = 74), an interaction signal (n = 76), or both (n = 4) (Figure 2). For the cases without any signal generated by the system, the lung medication mainly did not appear as actual medication at initiating the NS β-blocker (n = 94) or medication surveillance signals were not used in the pharmacy during the first dispensing (n = 20). For 23 patients, pharmacists could not recall the handling process of the first prescription, mainly due to the lack of digital archiving of handling the surveillance signals in the past. Processing of the medication surveillance signals is shown in Figure 3. In most cases, the patient was informed about the possibility of increased respiratory symptoms (n = 87).

| DISCUSSION
Our study showed that two-thirds of the interviewed prescribers of oral and ocular NS β-blockers prescribed this medication to asthma or COPD patients deliberately, considering the drug-disease interaction not to be relevant. During the interviews, ophthalmologists argued that they had never seen exacerbations in daily practice, which is not surprisingly as exacerbations are mainly treated by GPs or lung specialist. One-third of the prescribers was simply not aware of the presence of co-morbidity conditions.   The development of clinical guidelines only makes sense, if the care is actually implemented in health care practice. The use of guidelines in daily practice should be evaluated regularly, to identify room for improvement in practice and, if necessary, in the guidelines. 38 Our study showed that the implementation of the recommendation to avoid NS βblockers in patients with lung diseases had led to the general process of medication surveillance signals into the pharmacy information system, but that this computerized signals clearly did not necessarily result in actions by the pharmacists. In 39 of the 154 cases, the pharmacists ignored the signal and decided not to take any action (eg, consulting the prescriber or adjustment of the therapy). This is consistent with other research on decision support systems in pharmacies. 39 In almost half of the cases studied, pharmacists did not receive any medication surveillance signal in daily practice. Interaction signals were less generated for ocular-than for oral NS β-blockers ( switching of dose levels in the past, or on specific co-medication or co-morbidities. Evaluation of clinical outcomes is part of the implementation process, and this is an important topic for future research.
In conclusion, this study showed that, in contrast to prevailing guidelines, part of the interviewed prescribers did not assume NS βblockers to be contraindicated in the selected patients with asthma or COPD. There is no evidence to support this statement, and although the recommendation of avoiding NS β-blockers is implemented in the medication surveillance system in the pharmacy, pharmacists still need to be aware of the evidence-based clinical background of the generated signals and the importance of appropriate handling.
Further research is needed to evaluate to which extent the mentioned considerations are legitimate and to estimate clinical outcomes in patients with asthma and COPD, which are (deliberately or unintentionally) treated with NS β-blockers.